Skip to main content
. 2016 Feb 29;5:225. doi: 10.1186/s40064-016-1929-7

Table 3.

Descriptive statistics of personal, behavioural and social variables for the happy and unhappy groups

Variables (theoretical range) Happy
N = 400
Unhappy
N = 400
Statistical testa η2p
Mean (SD) Mean (SD)
Happiness (0–10) 10.0 (0) 2.5 (1.44)
Personal variables
Life satisfaction (0–10) 8.9 (1.77) 3.2 (2.08) F1,767 = 1455.86*** 0.66
Health (1–5) 4.1 (0.97) 2.9 (1.14) F1,767 = 181.23*** 0.19
Personal Wellbeing Index (0–10) 8.1 (1.38) 4.5 (1.82) F1,767 = 832.35*** 0.52
Behavioural variables
Leisure activities (1–8)
 Visiting cultural events 2.7 (0.97) 2.2 (1.06) F1,767 = 9.80** 0.02
 Active socializing and going out 2.9 (1.59) 2.3 (1.42) F1,767 = 2.92 ns 0.00
 Family and home activities 5.4 (0.93) 4.9 (1.15) F1,767 = 36.80*** 0.05
 Involvement in the community (1–3) 1.4 (0.64) 1.2 (0.53) F1, 767 = 3.64* 0.01
Social variables
National Wellbeing Index (0–10) 5.4 (2.08) 3.6 (1.71) F1,769 = 155.52*** 0.17
Trust in people (0–10) 5.0 (2.59) 3.5 (2.09) F1,769 = 75.67*** 0.09
Trust in the institutions (1–4)
Government institutions 1.9 (0.67) 1.7 (0.63) F1,769 = 19.33*** 0.03
Public institutions 2.5 (0.66) 2.2 (0.66) F1,769 = 51.13*** 0.06
Non-government institutions 2.3 (0.65) 2.1 (0.67) F1,769 = 19.18*** 0.02

ns non-significant

p < 0.05; ** p < 0.01; *** p < 0.001

aThree MACNOVAs were performed for personal, behavioural, and social variables separately to test the differences between happy and unhappy groups. The overall Wilks’s lambda for each MANCOVA was significant, p < 0.001. The tests of significance for individual dependent variables are presented. In all analyses the covariates were age, income and education level; η2p = Partial eta-squared, effect size measure